Compare commits

..

7 Commits

Author SHA1 Message Date
6cb611f2b5 test: deploy config validator tests (#690)
Some checks failed
Architecture Lint / Linter Tests (pull_request) Successful in 36s
Smoke Test / smoke (pull_request) Failing after 20s
Validate Config / YAML Lint (pull_request) Failing after 16s
Validate Config / JSON Validate (pull_request) Successful in 10s
Validate Config / Python Syntax & Import Check (pull_request) Failing after 1m15s
PR Checklist / pr-checklist (pull_request) Failing after 8m16s
Validate Config / Shell Script Lint (pull_request) Failing after 1m9s
Validate Config / Cron Syntax Check (pull_request) Successful in 18s
Validate Config / Deploy Script Dry Run (pull_request) Successful in 17s
Validate Config / Playbook Schema Validation (pull_request) Successful in 32s
Architecture Lint / Lint Repository (pull_request) Has been cancelled
Validate Config / Python Test Suite (pull_request) Has been cancelled
2026-04-15 03:23:06 +00:00
477fc6ce6f feat: sidecar config validation on deploy (#690) 2026-04-15 03:21:37 +00:00
d120526244 fix: add python3 shebang to scripts/visual_pr_reviewer.py (#681) 2026-04-15 02:57:53 +00:00
8596ff761b fix: add python3 shebang to scripts/diagram_meaning_extractor.py (#681) 2026-04-15 02:57:40 +00:00
7553fd4f3e fix: add python3 shebang to scripts/captcha_bypass_handler.py (#681) 2026-04-15 02:57:25 +00:00
71082fe06f fix: add python3 shebang to bin/soul_eval_gate.py (#681) 2026-04-15 02:57:14 +00:00
6d678e938e fix: add python3 shebang to bin/nostr-agent-demo.py (#681) 2026-04-15 02:57:00 +00:00
9 changed files with 483 additions and 221 deletions

View File

@@ -1,3 +1,4 @@
#!/usr/bin/env python3
"""
Full Nostr agent-to-agent communication demo - FINAL WORKING
"""

View File

@@ -1,3 +1,4 @@
#!/usr/bin/env python3
"""
Soul Eval Gate — The Conscience of the Training Pipeline

View File

@@ -1,3 +1,4 @@
#!/usr/bin/env python3
import json
from hermes_tools import browser_navigate, browser_vision

View File

@@ -0,0 +1,336 @@
#!/usr/bin/env python3
"""
deploy_config_validator.py — Pre-deploy config validation for timmy-config sidecar.
Validates YAML config before writing during deploy. Checks:
1. YAML syntax (pyyaml safe_load)
2. Required keys exist for the config type
3. Value types match expected schema
4. No banned providers referenced
5. Provider chain is well-formed
Usage:
# Validate a config file before deploy
python3 scripts/deploy_config_validator.py config.yaml
# Validate stdin (piped from deploy script)
cat config.yaml | python3 scripts/deploy_config_validator.py -
# Validate with expected type
python3 scripts/deploy_config_validator.py --type hermes config.yaml
# JSON output for CI/CD
python3 scripts/deploy_config_validator.py --json config.yaml
Exit codes:
0 — config is valid
1 — validation failed (errors printed to stderr)
2 — usage error
"""
import argparse
import json
import sys
from pathlib import Path
from typing import Any
try:
import yaml
except ImportError:
print("ERROR: PyYAML not installed. Run: pip install pyyaml", file=sys.stderr)
sys.exit(2)
# ── Schema Definitions ────────────────────────────────────────────────────────
# Required keys per config type
REQUIRED_KEYS = {
"hermes": {
"providers": {"type": list, "description": "List of provider configurations"},
},
"wizard": {
"providers": {"type": list, "description": "List of provider configurations"},
},
"ansible_inventory": {
"all": {"type": dict, "description": "Top-level inventory structure"},
},
"cron": {
"jobs": {"type": list, "description": "List of cron job definitions"},
},
"playbook": {
"name": {"type": str, "description": "Playbook name"},
},
"any": {}, # No required keys for generic validation
}
# Provider schema — each provider must have these keys
PROVIDER_REQUIRED = {"name", "model", "base_url"}
PROVIDER_ALLOWED_TYPES = {
"name": str,
"model": str,
"base_url": str,
"api_key_env": str,
"timeout": (int, float),
"reason": str,
}
# Banned provider patterns (from ansible inventory)
BANNED_PROVIDERS = {"anthropic", "claude"}
BANNED_MODEL_PATTERNS = ["claude-*", "anthropic/*", "*sonnet*", "*opus*", "*haiku*"]
# ── Validators ────────────────────────────────────────────────────────────────
class ValidationError:
def __init__(self, path: str, message: str, severity: str = "error"):
self.path = path
self.message = message
self.severity = severity
def __str__(self):
prefix = {"error": "ERROR", "warning": "WARN", "info": "INFO"}.get(self.severity, "???")
return f"[{prefix}] {self.path}: {self.message}"
def validate_yaml_syntax(text: str) -> tuple[Any | None, list[ValidationError]]:
"""Validate YAML syntax. Returns (parsed_data, errors)."""
errors = []
# Check for tabs
for i, line in enumerate(text.splitlines(), 1):
if "\t" in line:
errors.append(ValidationError(f"line {i}", "contains tab character (use spaces for YAML)", "warning"))
# Parse
try:
data = yaml.safe_load(text)
except yaml.YAMLError as e:
mark = getattr(e, "problem_mark", None)
if mark:
errors.append(ValidationError(
f"line {mark.line + 1}, col {mark.column + 1}",
f"YAML syntax error: {e.problem}"
))
else:
errors.append(ValidationError("(file)", f"YAML syntax error: {e}"))
return None, errors
if data is None:
errors.append(ValidationError("(file)", "empty or null config", "warning"))
return None, errors
return data, errors
def validate_required_keys(data: dict, config_type: str) -> list[ValidationError]:
"""Check that required keys exist."""
errors = []
schema = REQUIRED_KEYS.get(config_type, REQUIRED_KEYS["any"])
for key, spec in schema.items():
if key not in data:
errors.append(ValidationError(
f".{key}",
f"required key missing: {key} ({spec['description']})"
))
elif not isinstance(data[key], spec["type"]):
errors.append(ValidationError(
f".{key}",
f"expected {spec['type'].__name__}, got {type(data[key]).__name__}"
))
return errors
def validate_provider_chain(data: dict) -> list[ValidationError]:
"""Validate provider configurations."""
errors = []
providers = data.get("providers", [])
if not isinstance(providers, list):
return errors # Caught by required_keys check
for i, provider in enumerate(providers):
path = f".providers[{i}]"
if not isinstance(provider, dict):
errors.append(ValidationError(path, "provider must be a dict"))
continue
# Check required provider keys
for key in PROVIDER_REQUIRED:
if key not in provider:
errors.append(ValidationError(f"{path}.{key}", f"provider missing required key: {key}"))
elif not isinstance(provider[key], str):
errors.append(ValidationError(
f"{path}.{key}",
f"expected string, got {type(provider[key]).__name__}"
))
# Check for banned providers
name = provider.get("name", "").lower()
model = provider.get("model", "").lower()
for banned in BANNED_PROVIDERS:
if banned in name:
errors.append(ValidationError(
f"{path}.name",
f"banned provider: '{provider.get('name')}' (contains '{banned}')"
))
import fnmatch
for pattern in BANNED_MODEL_PATTERNS:
if fnmatch.fnmatch(model, pattern.lower()):
errors.append(ValidationError(
f"{path}.model",
f"banned model pattern: '{provider.get('model')}' matches '{pattern}'"
))
# Check value types
for key, val in provider.items():
expected = PROVIDER_ALLOWED_TYPES.get(key)
if expected and not isinstance(val, expected):
errors.append(ValidationError(
f"{path}.{key}",
f"expected {expected if isinstance(expected, type) else expected.__name__}, got {type(val).__name__}",
"warning"
))
# Check provider chain has at least one entry
if not providers:
errors.append(ValidationError(".providers", "provider chain is empty — no inference available"))
return errors
def validate_value_types(data: dict, path: str = "") -> list[ValidationError]:
"""Recursively check for obviously wrong value types."""
errors = []
if isinstance(data, dict):
for key, val in data.items():
full_path = f"{path}.{key}" if path else f".{key}"
# Ports should be integers
if key in ("port", "api_port", "hermes_port", "timeout") and val is not None:
if not isinstance(val, (int, float)):
errors.append(ValidationError(full_path, f"expected number, got {type(val).__name__}", "warning"))
# URLs should be strings starting with http
if key in ("base_url", "gitea_url", "url") and val is not None:
if isinstance(val, str) and not val.startswith(("http://", "https://")):
errors.append(ValidationError(full_path, f"URL should start with http:// or https://", "warning"))
# Recurse
errors.extend(validate_value_types(val, full_path))
elif isinstance(data, list):
for i, item in enumerate(data):
errors.extend(validate_value_types(item, f"{path}[{i}]"))
return errors
def validate_config(text: str, config_type: str = "any") -> list[ValidationError]:
"""Run all validations on a config text."""
# Step 1: YAML syntax
data, errors = validate_yaml_syntax(text)
if data is None:
return errors # Can't continue without parsed data
if not isinstance(data, dict):
if config_type != "any":
errors.append(ValidationError("(file)", f"expected dict for {config_type} config, got {type(data).__name__}"))
return errors
# Step 2: Required keys
errors.extend(validate_required_keys(data, config_type))
# Step 3: Provider chain validation (if providers exist)
if "providers" in data:
errors.extend(validate_provider_chain(data))
# Step 4: Value type checking
errors.extend(validate_value_types(data))
return errors
# ── Auto-detect config type ───────────────────────────────────────────────────
def detect_config_type(data: dict) -> str:
"""Guess config type from contents."""
if "providers" in data and "display" in data:
return "hermes"
if "providers" in data and "wizard_name" in data:
return "wizard"
if "all" in data and "children" in data.get("all", {}):
return "ansible_inventory"
if "jobs" in data:
return "cron"
if "name" in data and "hosts" in data:
return "playbook"
return "any"
# ── CLI ───────────────────────────────────────────────────────────────────────
def main():
parser = argparse.ArgumentParser(description="Pre-deploy config validation")
parser.add_argument("file", help="Config file to validate (use - for stdin)")
parser.add_argument("--type", choices=list(REQUIRED_KEYS.keys()),
help="Expected config type (auto-detected if omitted)")
parser.add_argument("--json", action="store_true", help="JSON output")
args = parser.parse_args()
# Read input
if args.file == "-":
text = sys.stdin.read()
filename = "<stdin>"
else:
path = Path(args.file)
if not path.exists():
print(f"ERROR: File not found: {path}", file=sys.stderr)
sys.exit(2)
text = path.read_text(encoding="utf-8", errors="replace")
filename = str(path)
# Detect type
config_type = args.type
if not config_type:
data, _ = validate_yaml_syntax(text)
if data and isinstance(data, dict):
config_type = detect_config_type(data)
else:
config_type = "any"
# Validate
errors = validate_config(text, config_type)
# Output
if args.json:
result = {
"file": filename,
"type": config_type,
"valid": not any(e.severity == "error" for e in errors),
"error_count": sum(1 for e in errors if e.severity == "error"),
"warning_count": sum(1 for e in errors if e.severity == "warning"),
"errors": [{"path": e.path, "message": e.message, "severity": e.severity} for e in errors],
}
print(json.dumps(result, indent=2))
else:
if errors:
print(f"Config validation FAILED: {filename} (type: {config_type})", file=sys.stderr)
for e in errors:
print(f" {e}", file=sys.stderr)
else:
print(f"Config validation PASSED: {filename} (type: {config_type})")
# Exit code
if any(e.severity == "error" for e in errors):
sys.exit(1)
sys.exit(0)
if __name__ == "__main__":
main()

View File

@@ -1,3 +1,4 @@
#!/usr/bin/env python3
import json
from hermes_tools import browser_navigate, browser_vision

View File

@@ -1,176 +0,0 @@
#!/usr/bin/env python3
"""PR Triage Automation -- Categorize, deduplicate, report (#659)."""
import argparse, json, os, re, sys, subprocess
from collections import Counter, defaultdict
from datetime import datetime
from urllib.request import Request, urlopen
from urllib.error import HTTPError
def _token():
t = os.environ.get("GITEA_TOKEN", "")
if not t:
p = os.path.expanduser("~/.config/gitea/token")
if os.path.exists(p):
t = open(p).read().strip()
return t
def _api(url, token, method="GET", data=None):
h = {"Authorization": "token " + token, "Accept": "application/json"}
body = json.dumps(data).encode() if data else None
if data:
h["Content-Type"] = "application/json"
req = Request(url, data=body, headers=h, method=method)
try:
return json.loads(urlopen(req, timeout=30).read())
except HTTPError:
return None
def fetch_prs(base, token, owner, repo):
prs, page = [], 1
while True:
b = _api(base + "/api/v1/repos/" + owner + "/" + repo + "/pulls?state=open&limit=50&page=" + str(page), token)
if not b:
break
prs.extend(b)
if len(b) < 50:
break
page += 1
return prs
def fetch_issues(base, token, owner, repo):
iss, page = {}, 1
while True:
b = _api(base + "/api/v1/repos/" + owner + "/" + repo + "/issues?state=open&limit=50&page=" + str(page), token)
if not b:
break
for i in b:
if "pull_request" not in i:
iss[i["number"]] = i
if len(b) < 50:
break
page += 1
return iss
def categorize(pr):
c = (pr.get("title", "") + " " + pr.get("body", "") + " " + " ".join(l.get("name", "") for l in pr.get("labels", []))).lower()
for kw, cat in [("training data", "training-data"), ("dpo", "training-data"), ("grpo", "training-data"),
("fix:", "bug-fix"), ("bug", "bug-fix"), ("hotfix", "bug-fix"),
("feat:", "feature"), ("feature", "feature"),
("refactor", "maintenance"), ("cleanup", "maintenance"),
("doc", "documentation"), ("test", "testing"), ("infra", "infrastructure")]:
if kw in c:
return cat
return "other"
def refs(pr):
return [int(m) for m in re.findall(r"#(\d+)", pr.get("title", "") + " " + pr.get("body", ""))]
def find_duplicates(prs):
by = defaultdict(list)
for p in prs:
for r in refs(p):
by[r].append(p)
return [g for g in by.values() if len(g) > 1]
def health(pr, issues):
r = refs(pr)
created = datetime.fromisoformat(pr["created_at"].replace("Z", "+00:00"))
updated = datetime.fromisoformat(pr["updated_at"].replace("Z", "+00:00"))
now = datetime.now(created.tzinfo)
return {
"pr": pr["number"], "title": pr["title"], "head": pr["head"]["ref"],
"category": categorize(pr), "refs": r,
"open": [x for x in r if x in issues], "closed": [x for x in r if x not in issues],
"age": (now - created).days, "stale": (now - updated).days,
"mergeable": pr.get("mergeable"), "author": pr.get("user", {}).get("login", ""),
}
def report(repo, checks, dups):
lines = ["# PR Triage -- " + repo,
"Generated: " + datetime.now().strftime("%Y-%m-%d %H:%M"),
"Open PRs: " + str(len(checks)), "", "## Summary", ""]
cats = Counter(h["category"] for h in checks)
lines.append("| Category | Count |")
lines.append("|----------|-------|")
for c, n in cats.most_common():
lines.append("| " + c + " | " + str(n) + " |")
stale = [h for h in checks if h["stale"] > 7]
lines.extend(["", "Stale (>7d): " + str(len(stale)),
"Duplicate groups: " + str(len(dups)), ""])
if dups:
lines.append("## Duplicates")
for g in dups:
rs = set()
for p in g:
rs.update(refs(p))
lines.append("Issues " + ", ".join("#" + str(r) for r in sorted(rs)) + ":")
for p in g:
lines.append(" - #" + str(p["number"]) + ": " + p["title"])
lines.append("")
if stale:
lines.append("## Stale (>7d)")
for h in sorted(stale, key=lambda x: x["stale"], reverse=True):
lines.append("- #" + str(h["pr"]) + ": " + h["title"] + " -- " + str(h["stale"]) + "d")
lines.append("")
lines.append("## All PRs")
lines.append("| # | Title | Category | Age | Stale | Merge |")
lines.append("|---|-------|----------|-----|-------|-------|")
for h in sorted(checks, key=lambda x: x["pr"]):
m = "Y" if h["mergeable"] else ("N" if h["mergeable"] is False else "?")
s = str(h["stale"]) + "d" if h["stale"] > 7 else "-"
lines.append("| " + str(h["pr"]) + " | " + h["title"][:50] + " | " + h["category"] +
" | " + str(h["age"]) + "d | " + s + " | " + m + " |")
return chr(10).join(lines)
def main():
p = argparse.ArgumentParser(description="PR Triage Automation")
p.add_argument("--base-url", default="https://forge.alexanderwhitestone.com")
p.add_argument("--owner", default="Timmy_Foundation")
p.add_argument("--repo", default="")
p.add_argument("--json", action="store_true", dest="js")
p.add_argument("--output", default="")
a = p.parse_args()
token = _token()
if not token:
print("No token"); sys.exit(1)
repo = a.repo
if not repo:
try:
remote = subprocess.check_output(["git", "remote", "get-url", "origin"], text=True).strip()
m = re.search(r"[/:](\w[\w-]*)/(\w[\w-]*?)(?:\.git)?$", remote)
if m:
a.owner, repo = m.group(1), m.group(2)
except Exception:
pass
if not repo:
print("No repo specified"); sys.exit(1)
print("Triaging " + a.owner + "/" + repo + "...", file=sys.stderr)
prs = fetch_prs(a.base_url, token, a.owner, repo)
issues = fetch_issues(a.base_url, token, a.owner, repo)
checks = [health(pr, issues) for pr in prs]
dups = find_duplicates(prs)
if a.js:
print(json.dumps({"repo": repo, "prs": checks,
"duplicates": [[{"number": p["number"], "title": p["title"]} for p in g] for g in dups]},
indent=2))
else:
r = report(repo, checks, dups)
print(r)
if a.output:
with open(a.output, "w") as f:
f.write(r)
print("\n" + str(len(checks)) + " PRs, " + str(len(dups)) + " duplicate groups", file=sys.stderr)
if __name__ == "__main__":
main()

View File

@@ -1,3 +1,4 @@
#!/usr/bin/env python3
import json
from hermes_tools import browser_navigate, browser_vision

View File

@@ -0,0 +1,142 @@
#!/usr/bin/env python3
"""Tests for deploy_config_validator.py"""
import json
import sys
import os
import pytest
sys.path.insert(0, os.path.dirname(os.path.dirname(__file__)))
from scripts.deploy_config_validator import (
validate_yaml_syntax,
validate_required_keys,
validate_provider_chain,
validate_value_types,
validate_config,
detect_config_type,
ValidationError,
)
class TestYAMLSyntax:
def test_valid_yaml(self):
data, errors = validate_yaml_syntax("key: value\nlist:\n - a\n - b\n")
assert data is not None
assert len(errors) == 0
def test_invalid_yaml(self):
data, errors = validate_yaml_syntax("key: [unclosed")
assert data is None
assert len(errors) > 0
def test_empty_yaml(self):
data, errors = validate_yaml_syntax("")
assert data is None
assert any("empty" in e.message for e in errors)
def test_tabs_warning(self):
data, errors = validate_yaml_syntax("key:\tvalue\n")
assert any("tab" in e.message for e in errors)
class TestRequiredKeys:
def test_missing_key(self):
errors = validate_required_keys({}, "hermes")
assert any("providers" in e.message for e in errors)
def test_wrong_type(self):
errors = validate_required_keys({"providers": "not-a-list"}, "hermes")
assert any("expected list" in e.message for e in errors)
def test_valid(self):
errors = validate_required_keys({"providers": []}, "hermes")
provider_errors = [e for e in errors if "providers" in e.message and "missing" in e.message]
assert len(provider_errors) == 0
class TestProviderChain:
def test_empty_providers(self):
errors = validate_provider_chain({"providers": []})
assert any("empty" in e.message for e in errors)
def test_missing_name(self):
errors = validate_provider_chain({"providers": [{"model": "test", "base_url": "http://x"}]})
assert any("name" in e.message and "missing" in e.message for e in errors)
def test_banned_provider(self):
errors = validate_provider_chain({"providers": [
{"name": "anthropic", "model": "claude-3", "base_url": "http://x"}
]})
assert any("banned provider" in e.message for e in errors)
def test_banned_model(self):
errors = validate_provider_chain({"providers": [
{"name": "test", "model": "claude-sonnet-4", "base_url": "http://x"}
]})
assert any("banned model" in e.message for e in errors)
def test_valid_providers(self):
errors = validate_provider_chain({"providers": [
{"name": "kimi-coding", "model": "kimi-k2.5", "base_url": "https://api.kimi.com/v1"}
]})
provider_errors = [e for e in errors if e.severity == "error"]
assert len(provider_errors) == 0
class TestValueTypes:
def test_string_port(self):
errors = validate_value_types({"port": "8080"})
assert any("port" in e.path and "number" in e.message for e in errors)
def test_valid_port(self):
errors = validate_value_types({"port": 8080})
port_errors = [e for e in errors if "port" in e.path]
assert len(port_errors) == 0
def test_bad_url(self):
errors = validate_value_types({"base_url": "not-a-url"})
assert any("URL" in e.message for e in errors)
class TestDetectConfigType:
def test_hermes(self):
t = detect_config_type({"providers": [], "display": {}})
assert t == "hermes"
def test_ansible(self):
t = detect_config_type({"all": {"children": {"wizards": {}}}})
assert t == "ansible_inventory"
def test_unknown(self):
t = detect_config_type({"random": "data"})
assert t == "any"
class TestFullValidation:
def test_valid_hermes_config(self):
text = """
providers:
- name: kimi-coding
model: kimi-k2.5
base_url: https://api.kimi.com/coding/v1
timeout: 120
display:
skin: default
"""
errors = validate_config(text, "hermes")
assert not any(e.severity == "error" for e in errors)
def test_banned_provider_catches(self):
text = """
providers:
- name: anthropic
model: claude-sonnet-4
base_url: https://api.anthropic.com
"""
errors = validate_config(text, "hermes")
assert any("banned" in e.message for e in errors)
def test_missing_providers(self):
text = "display:\n skin: default\n"
errors = validate_config(text, "hermes")
assert any("providers" in e.message and "missing" in e.message for e in errors)

View File

@@ -1,45 +0,0 @@
"""Tests for PR triage automation (#659)."""
import pytest
class TestCategorize:
def _pr(self, title="", body=""):
return {"title": title, "body": body, "labels": []}
def test_training(self):
from scripts.pr_triage import categorize
assert categorize(self._pr("Add DPO pairs")) == "training-data"
def test_bug(self):
from scripts.pr_triage import categorize
assert categorize(self._pr("fix: crash")) == "bug-fix"
def test_feature(self):
from scripts.pr_triage import categorize
assert categorize(self._pr("feat: dark mode")) == "feature"
def test_other(self):
from scripts.pr_triage import categorize
assert categorize(self._pr("random")) == "other"
class TestRefs:
def test_simple(self):
from scripts.pr_triage import refs
assert 123 in refs({"title": "Fix #123", "body": ""})
def test_multiple(self):
from scripts.pr_triage import refs
r = refs({"title": "", "body": "Closes #100, Refs #200"})
assert 100 in r and 200 in r
class TestDuplicates:
def test_found(self):
from scripts.pr_triage import find_duplicates
prs = [{"title": "", "body": "Fix #1", "number": 1, "head": {"ref": "a"}, "created_at": "2026-01-01T00:00:00Z", "updated_at": "2026-01-01T00:00:00Z", "user": {}},
{"title": "", "body": "Refs #1", "number": 2, "head": {"ref": "b"}, "created_at": "2026-01-01T00:00:00Z", "updated_at": "2026-01-01T00:00:00Z", "user": {}}]
assert len(find_duplicates(prs)) == 1
def test_none(self):
from scripts.pr_triage import find_duplicates
prs = [{"title": "", "body": "Fix #1", "number": 1, "head": {"ref": "a"}, "created_at": "2026-01-01T00:00:00Z", "updated_at": "2026-01-01T00:00:00Z", "user": {}},
{"title": "", "body": "Fix #2", "number": 2, "head": {"ref": "b"}, "created_at": "2026-01-01T00:00:00Z", "updated_at": "2026-01-01T00:00:00Z", "user": {}}]
assert find_duplicates(prs) == []